Higher cycling costs loom as solar, wind resources grow

Latest reports from one consulting firm, with arguably more experience than anyone else in estimating the costs of cycling fossil units, suggest this: Gas-fired units may be in for elevated levels of pain, if not a death spiral, trying to make ends meet as more renewable resources come online.

This shouldn’t be too surprising to owner/operators. After all, virtually every facility has experienced changes to capacity factor, operating tempo, starts and stops, and load ramping as renewables have grown in a given independent system operator (ISO) service territory, or region. The coming conundrum is that capacity factors and electricity prices are expected to decrease, cutting into revenues, as even greater cycling and ramping impose more cost on the facility. At the same time, aging gas-fired units are more vulnerable to escalating equivalent forced outage rates (EFOR) and risk of protracted outages from significant failures rooted in cycling impacts from years ago.

The analysis presented here is mostly to help executives make data-driven risk management investment decisions. But individual plant managers and specialists need to assess their vulnerabilities within a broader context. Estimating cycling impacts and cost may be more art than science, but few would argue that the consequences can be catastrophic.

PJM focus. The fossil generator’s pain in providing the flexibility “behind the scenes” is the grid operator’s glory in accommodating higher and higher levels of renewable energy. The former is often the implicit message while the latter is the explicit one for the public.

Capital costs of solar and wind keep declining. Since there is no fuel cost, the marginal cost begins to approach zero, asymptotically anyway. Overall production costs to the grid are lower. Electricity rates remain stable or decline. What’s not to like?

However, each megawatt-hour of flexibility from the gas-fired facility has an increasing incremental cost of cycling attached to it.

A recent monster of a study for PJM illuminates all of this. The PJM Renewable Integration Study, a 2014 report led by General Electric International Inc, assessed system impacts of increased penetration of wind and solar resources on operation of the PJM grid. Scenarios of up to 30% renewables by 2026 were investigated.

The report’s headline conclusion is this: “PJM, with adequate transmission expansion and additional regulating reserves, will not have any significant issues operating with up to 30% of its energy provided by wind and solar generation.”

Here’s the critical sub-conclusion for gas-fired asset owner/operators: “Every scenario examined resulted in lower fuel and variable operations and maintenance (O&M) costs, as well as lower average locational marginal prices (LMP). The lower LMPs, when combined with reduced capacity factors, resulted in lower gross and net revenues for the conventional generation resources.”

The report goes on to note that the increase in variable O&M costs from added starts, stops, and ramping of conventional units is “small relative to the value of the fuel displacement,” and did not significantly affect the overall impact of renewables generation (Fig 1).

That’s cold comfort to fossil-fired plant owners and operators as any added costs in a declining revenue environment are bad for survival. In addition, conventional generators may have to compete with emerging grid-scale storage facilities for some of that flexibility, especially in the short-duration frequency regulation market. And what happens if gas prices begin to rise? Ominously, gas-fired combined cycles were shown to have the greatest change in cycling damage compared to other conventional fossil resources (Fig 2).

One of the recommendations included in the report is that PJM should explore the reasons for ramping constraints on specific units, and determine whether the limitations are technical, contractual, or otherwise, then investigate possible methods of improving ramp-rate performance.

It’s easy to envision other potential consequences and solutions:

      • PJM could add monetary incentives, such as capacity-type payments, so conventional resources can continue to serve.

      • Conventional resources could be consolidated into a few hands to control prices bid into the PJM market under flexible-resource operating modes.

Although the report’s principal investigator is GE, Intertek AIM Engineering Consulting, Santa Clara, Calif, provided the analysis on the impact of cycling on variable O&M costs. Intertek acquired Aptech Engineering Services in 2009; CCJ editors are familiar with Aptech’s cost of cycling models and assessments dating back to 1986.

Similar stories abound. Some version of the PJM scenarios is unfolding in most other markets and regions around the country. In a presentation, “Impact of cycling on availability,” at the 2017 501F Users Group conference, Intertek’s Nikhil Kumar analyzed three markets from different parts of the country: California, Iowa, and Texas.

He presented graphs on national trends, including the rise of gas and the fall of coal over the last 15 years, growth in the average number of starts per year and in average annual operating hours for peaking and combined-cycle units, and relative increases in hot, warm, and cold starts. These trends are familiar to most CCJ readers.

Collective unit operating data in the three markets studied also were presented, all of which validate quantitatively what readers know anecdotally: gas-fired units cycle more as renewable penetration rates grow. One interesting feature of the data is that conventional units struggle more to follow the wind in Iowa because the resource is much stronger than, say, in Texas. Fossil units in Texas have experienced less cycling following renewables than in Iowa or California.

The researchers suggest that accumulating cycling damage be measured not only in equivalent operating hours but in equivalent hot starts, the latter meaning that each cold and warm start imposes the equivalent of multiple hot starts on the unit in terms of thermally induced cyclic stress, fatigue damage, strain softening, and other mechanical properties—especially in thick-walled components.

The most important point perhaps is that combined-cycle owner/operators need to take a page from recent history and understand how EFORs rise as units age (Fig 3), and the direct relationship to total annual equivalent operating cycles. While the trend lines are clear, the actual damage mechanisms are difficult to identify. For combined cycles, each hot, warm, and cold start has an incremental impact on the EFOR (Fig 4), and a cost associated with it.

   

Bottom-line impacts. In a follow-up interview, Kumar noted that one of the most important takeaways is that new combined cycles must be designed for greater ramping and cycling flexibility. Despite the flexible unit designs being offered by the OEMs, Kumar argues that combined cycles are still being built without the requisite features for deep cycling and quick ramping. Utilities displacing coal capacity with new gas capacity are especially prone to short-sighted baseload operations design.

As an example, Kumar points to a new plant designed with an HRSG reheater drain line which makes a 180 deg turn and then flows straight up before reaching an HRSG blowdown/drain tank (Fig 5). The tank inlets are located 12-15 ft above the grade of the lower HRSG headers. Such designs allow water to accumulate at low points which then lead to damaging temperature gradients on thick-walled pressure parts under frequent cycling.

Many combined cycles built during the now famed bubble years of 1997-2003 will experience a double whammy of sorts, Kumar stresses. First, the vast majority of these plants were designed for baseload operation during a period of intense competition, queues ordering major components from primary vendors, and cutthroat bidding among project developers vying for long-term power purchase agreements as independent and merchant suppliers. These plants often lack one or more of the following:

      • Automated drain systems.

      • Auxiliary boiler to raise steam during startup.

      • Stack damper.

      • Steam bypass system.

      • Nitrogen or inert blanketing on the condensate storage tank to prevent oxygen ingress.

      • Condenser steam dump.

      • HRSG drain thermocouples.

      • Reheater attemperator outlet thermocouples.

      • HRSG stress indicator.

      • Tight control and monitoring of feedwater chemistry.

      • Steam sparging for reheat and superheat steam temperature control.

      • Sophisticated inspection and monitoring capabilities.

The second “whammy” is that these units are hitting the 15-20-yr point in the aging cycle. Intertek’s data covering hundreds of units over decades of operation clearly shows that such units are more vulnerable to escalating EFOR and, importantly, high-impact outage events which can force a unit out of service for months.

One vintage 1995 combined cycle Kumar mentioned, operating in the Pacific Northwest, suffered a high-cycle fatigue failure from a crack that originated at the root of a corrosion pit in the steam turbine rotor, ultimately forcing the unit out of service for five months. Official testimony from the owner/operator revealed that “the fact that the rotor has not failed completely in spite of the extensive cracking suggest that the conditions observed were not recent events but had been in existence for some time.” In other words, the damage accumulated over long periods of time.

Ultimately, Kumar says, grid operators and plant owners will have to make some tough decisions about how to retain the flexibility necessary to accommodate ever increasing renewables penetration. For plants, the decisions will depend on whether they are regulated, merchant, or under long-term PPAs and the degree of cycling and on/off operation they are expected to accommodate.

One challenge to making the requisite investments is gas-fired plants change owners at a dizzying pace. Who can justify investments for operating flexibility under short-term ownership horizons?

Grid operators will have to adapt and adopt mechanisms for conventional fossil resources to be properly compensated for their flexibility. Most ISOs already have various financial products to encourage flexible operation—such as frequency-regulation payments, capacity payments, day-ahead bidding, ramping products, resource adequacy payments, etc. Another one Kumar thinks should be considered is a special payment for operating at ultra-low load for extended periods.

Like most things in life, grid operation with 30% renewable penetration may not, on paper, pose significant issues from the perspective of the grid operator, but only if in the real-world facilities providing the necessary flexibility, hidden from public view, are appropriately compensated.

Posted in WTUI |

Comments are closed.